Skip to content

Add token streaming support for XMLAdapter #8478

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
+97 −4
Open

Add token streaming support for XMLAdapter #8478

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
84f2c45
Add XMLAdapter support to StreamListener
chenmoneygithub Jul 1, 2025
124ad8a
Add XMLAdapter streaming and tests
chenmoneygithub Jul 1, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 20 additions & 2 deletions dspy/streaming/streaming_listener.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@

from dspy.adapters.chat_adapter import ChatAdapter
from dspy.adapters.json_adapter import JSONAdapter
from dspy.adapters.xml_adapter import XMLAdapter
from dspy.dsp.utils.settings import settings
from dspy.streaming.messages import StreamResponse

Expand Down Expand Up @@ -51,6 +52,9 @@ def __init__(
self.chat_adapter_start_identifier = f"[[ ## {self.signature_field_name} ## ]]"
self.chat_adapter_end_identifier = re.compile(r"\[\[ ## (\w+) ## \]\]")

self.xml_adapter_start_identifier = f"<{self.signature_field_name}>"
self.xml_adapter_end_identifier = re.compile(rf"</{self.signature_field_name}>")

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Since we already have three adapters, shall we define a mapping for adapter-specific parameters? Alternatively, we can move the identifier into each adapter class for better encapsulation

def _buffered_message_end_with_start_identifier(self, concat_message: str, start_identifier: str) -> str:
for i in range(len(concat_message)):
if start_identifier.startswith(concat_message[len(concat_message) - i - 1 :]):
Expand All @@ -63,14 +67,19 @@ def receive(self, chunk: ModelResponseStream):
end_identifier = self.json_adapter_end_identifier

start_indicator = '"'
elif isinstance(settings.adapter, XMLAdapter):
start_identifier = self.xml_adapter_start_identifier
end_identifier = self.xml_adapter_end_identifier

start_indicator = "<"
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we also make start_indicator an instance variable like we do for start_identifier and end_identifier?

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

As I mentioned above, maybe it's better to move the logic into each adapter class and remove if-else branch

start_identifier = settings.adapter.get_start_identifier(self.signature_field_name)
end_identifier= settings.adapter.get_end_identifier(self.signature_field_name)
start_indicator= settings.adapter.get_start_indicator()

elif isinstance(settings.adapter, ChatAdapter) or settings.adapter is None:
start_identifier = self.chat_adapter_start_identifier
end_identifier = self.chat_adapter_end_identifier

start_indicator = "["
else:
raise ValueError(
f"Unsupported adapter for streaming: {settings.adapter}, please use either ChatAdapter or "
f"Unsupported adapter for streaming: {settings.adapter}, please use either ChatAdapter, XMLAdapter or "
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: Can we define const for a list of adapters that support streaming?

"JSONAdapter for streaming purposes."
)

Expand Down Expand Up @@ -175,12 +184,21 @@ def flush(self) -> str:
else:
boundary_index = len(last_tokens)
return last_tokens[:boundary_index]
elif isinstance(settings.adapter, XMLAdapter):
boundary_index = last_tokens.find(f"</{self.signature_field_name}>")
if boundary_index == -1:
boundary_index = len(last_tokens)
return last_tokens[:boundary_index]
elif isinstance(settings.adapter, ChatAdapter) or settings.adapter is None:
boundary_index = last_tokens.find("[[")
return last_tokens[:boundary_index]
boundary_index = last_tokens.find(f"</{self.signature_field_name}>")
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

q: why do we need this change for ChatAdapter?

if boundary_index == -1:
boundary_index = len(last_tokens)
return last_tokens[:boundary_index]
else:
raise ValueError(
f"Unsupported adapter for streaming: {settings.adapter}, please use either ChatAdapter or "
f"Unsupported adapter for streaming: {settings.adapter}, please use either ChatAdapter, XMLAdapter or "
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ditto

"JSONAdapter for streaming purposes."
)

Expand Down
79 changes: 77 additions & 2 deletions tests/streaming/test_streaming.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -388,7 +388,7 @@ async def gpt_4o_mini_stream_2():
async def completion_side_effect(*args, **kwargs):
return stream_generators.pop(0)() # return new async generator instance

with mock.patch("litellm.acompletion", side_effect=completion_side_effect) as mock_completion:
with mock.patch("litellm.acompletion", side_effect=completion_side_effect):
program = dspy.streamify(
MyProgram(),
stream_listeners=[
Expand Down Expand Up @@ -484,7 +484,7 @@ async def gpt_4o_mini_stream_2(*args, **kwargs):

with mock.patch(
"litellm.acompletion", new_callable=AsyncMock, side_effect=[gpt_4o_mini_stream_1(), gpt_4o_mini_stream_2()]
) as mock_completion:
):
program = dspy.streamify(
MyProgram(),
stream_listeners=[
Expand Down Expand Up @@ -763,3 +763,78 @@ async def completion_side_effect(*args, **kwargs):
concat_message = "".join([chunk.chunk for chunk in all_chunks])
# The listener functions twice.
assert concat_message == "To get to the other side!To get to the other side!"

@pytest.mark.anyio
async def test_stream_listener_returns_correct_chunk_xml_adapter():
class MyProgram(dspy.Module):
def __init__(self):
super().__init__()
self.predict1 = dspy.Predict("question->answer")
self.predict2 = dspy.Predict("question,answer->judgement")

def forward(self, question, **kwargs):
answer = self.predict1(question=question, **kwargs).answer
judgement = self.predict2(question=question, answer=answer, **kwargs)
return judgement

async def xml_stream_1(*args, **kwargs):
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="<"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="answer"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=">"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="To"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=" get"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=" to"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=" the"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=" other"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=" side"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="!"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="<"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="/answer"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=">"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="<"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="completed"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=">"))])

async def xml_stream_2(*args, **kwargs):
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="<"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="judgement"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=">"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="The"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=" answer"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=" is"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=" humorous"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="."))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="<"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="/judgement"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=">"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="<"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content="completed"))])
yield ModelResponseStream(model="gpt-4o-mini", choices=[StreamingChoices(delta=Delta(content=">"))])

stream_generators = [xml_stream_1, xml_stream_2]

async def completion_side_effect(*args, **kwargs):
return stream_generators.pop(0)()

with mock.patch("litellm.acompletion", side_effect=completion_side_effect):
program = dspy.streamify(
MyProgram(),
stream_listeners=[
dspy.streaming.StreamListener(signature_field_name="answer"),
dspy.streaming.StreamListener(signature_field_name="judgement"),
],
)
with dspy.context(lm=dspy.LM("openai/gpt-4o-mini", cache=False), adapter=dspy.XMLAdapter()):
output = program(question="why did a chicken cross the kitchen?")
all_chunks = []
async for value in output:
if isinstance(value, dspy.streaming.StreamResponse):
all_chunks.append(value)

assert all_chunks[0].predict_name == "predict1"
assert all_chunks[0].signature_field_name == "answer"
assert all_chunks[0].chunk == "To get to the other side!"

assert all_chunks[1].predict_name == "predict2"
assert all_chunks[1].signature_field_name == "judgement"
assert all_chunks[1].chunk == "The answer is humorous."